The boll weevil, Anthonomus grandis (Boheman), has been a major insect pest of cotton production in the US, accounting for yield losses and control costs on the order of several billion US dollars since the introduction of the pest in 1892. Boll weevil eradication programs have eliminated reproducing populations in nearly 94%, and progressed toward eradication within the remaining 6%, of cotton production areas. However, the ability of weevils to disperse and reinfest eradicated zones threatens to undermine the previous investment toward eradication of this pest. In this study, the HYSPLIT atmospheric dispersion model was used to simulate daily wind-aided dispersal of weevils from the Lower Rio Grande Valley (LRGV) of southern Texas and northeastern Mexico. Simulated weevil dispersal was compared with weekly capture of weevils in pheromone traps along highway trap lines between the LRGV and the South Texas/Winter Garden zone of the Texas Boll Weevil Eradication Program. A logistic regression model was fit to the probability of capturing at least one weevil in individual pheromone traps relative to specific values of simulated weevil dispersal, which resulted in 60.4% concordance, 21.3% discordance, and 18.3% ties in estimating captures and non-captures. During the first full year of active eradication with widespread insecticide applications in 2006, the dispersal model accurately estimated 71.8%, erroneously estimated 12.5%, and tied 15.7% of capture and non-capture events. Model simulations provide a temporal risk assessment over large areas of weevil reinfestation resulting from dispersal by prevailing winds. Eradication program managers can use the model risk assessment information to effectively schedule and target enhanced trapping, crop scouting, and insecticide applications.